嵌入式系统低功耗优化技术在肺音信号处理中的应用

发布时间：2018-09-13 06:59

【摘要】：降低嵌入式系统的功耗有助于提高其稳定性、延长电池的续航时间、拓展产品的适用范围、减小其尺寸和重量。在全球倡导发展绿色经济的大背景下，降低嵌入式系统的功耗也是节能减排、保护生态环境的必然要求。本文以一款用于监测麻醉术中病人呼吸功能的嵌入式系统为研究对象，对它的低功耗优化问题进行深入探讨，以期降低其功耗，延长电池的使用时间，提高该系统的稳定性。首先，简要介绍嵌入式系统低功耗优化技术的研究背景和研究现状，在此基础上，详细阐述了CMOS电路的功耗特性、DVS和DPM技术、源程序级和算法级软件低功耗优化的关键技术及其相关原理。对MPU、存储系统、外围设备的低功耗优化的一般性方法进行了研究和总结。提出了一种可降低存储系统功耗的RAM自适应可重构动态分配机制，并给出了相应算法。其次，根据肺音信号处理系统的结构、功能和特点，着重研究该系统低功耗优化的策略与方法。提出可通过动态变换DSP数字信号处理器的运行速度、使用低功耗指令、正确处理未用引脚等措施来降低DSP的功耗；可对该系统的软件部分进行低功耗优化，通过充分利用片内存储空间和寄存器、合理使用存储器的低功耗模式、恰当地使用内联函数和宏、增强程序的局部性、相关数据的连续存储、循环体的适度展开与合并等策略与方法的联合使用来降低存储系统的功耗；对于LCD等外设，可通过使用中断方式、外设的低功耗模式、降低服务质量等方法来降低它们的功耗。最后，通过三组实验来检验这些优化方法的有效性，并对实验结果作出分析和评估。结果表明，优化后的肺音信号处理系统在工作状态、待机状态及模拟用户使用时，其平均功耗均有不同程度的降低，从而实现了降低该系统的功耗的目标。
[Abstract]:Reducing the power consumption of the embedded system can improve its stability, prolong the battery life, expand the application range of the product, and reduce its size and weight. In the context of global advocacy for the development of a green economy, reducing the power consumption of embedded systems is also an inevitable requirement for energy conservation and emission reduction, as well as for the protection of the ecological environment. In this paper, an embedded system for monitoring patient respiratory function in anesthesiology is studied, and its low power consumption optimization is discussed in order to reduce the power consumption, prolong the battery life and improve the stability of the system. First of all, the research background and current situation of low power optimization technology in embedded system are briefly introduced. On this basis, the power characteristics of CMOS circuit are described in detail, including DVS and DPM technology. The key techniques and related principles of low power optimization for source and algorithmic software. In this paper, the general methods of low power optimization for MPU, storage systems and peripherals are studied and summarized. A RAM adaptive reconfigurable dynamic allocation mechanism is proposed, which can reduce the power consumption of storage system, and the corresponding algorithm is given. Secondly, according to the structure, function and characteristics of lung sound signal processing system, the strategy and method of low power consumption optimization are studied. It is proposed that the power consumption of DSP can be reduced by dynamically changing the speed of DSP digital signal processor, using low power instruction and correctly handling unused pins, and the software part of the system can be optimized with low power consumption. By making full use of the on-chip storage space and registers, using the low power mode of memory reasonably, using inline functions and macros properly, enhancing the locality of the program and the continuous storage of related data, The combination of strategies and methods such as appropriate expansion and merging of loop bodies can reduce the power consumption of storage systems. For peripheral devices such as LCD, the power consumption can be reduced by using interrupt mode, low power mode of peripheral devices, and reducing quality of service. Finally, the effectiveness of these optimization methods is tested by three groups of experiments, and the experimental results are analyzed and evaluated. The results show that the average power consumption of the optimized lung sound signal processing system is reduced in different degrees in the working state, standby state and analogue user, thus achieving the goal of reducing the power consumption of the system.
【学位授予单位】：哈尔滨理工大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN911.7;TP368.1

【参考文献】